Semantic Data Control

“——————————————————————————————————————————————————



Support translation: http://amzn.to/1Z7d5oc
——————————————————————————————————————————————————

Semantic Data Control

Semantic data control typically includes view management, security control and semantic integrity control. These functions ensure that authorised users perform correct operations on the database. This contributes to database integrity.

Views in Distributed DBMS

A view in a distributed database is derived from fragmented relations stored at different sites. When a view is defined, its name and its retrieval query are stored in a catalogue.

The mapping of a query expressed on views is done in the same way as in centralised systems. In centralized systems it is done through query modification. A distributed query is processed by the distributed query processor.

Views in a distributed DBMS can be done by using rules similar to fragment definition rules. The query processor maps the distributed query into a query on physical fragments.

Views derived from the distributed database may be costly to evaluate. View derivation is done by merging the view qualification with the query qualification. An alternative is to use snapshots. A snapshot represents a particular state of a database. This means that it is static. Snapshots are useful when users are not interested in seeing the most recent version of the database. Access through snapshots is adequate for queries that are predefined.

Distributed Authorization Control

More complex than in central databases due to the fact that objects and users can be distributed. These problems are:
– Remote user authentication
– Management of distributed authorization rules
– Handling of views and user groups.

To prevent remote user access by unauthorized users, 2 approaches are used:
– Information of authorized users are stored on all sites in the catalogue (more costly)
– All sites of distributed DBMS authenticate themselves like users do. (necessary if user information is not replicated).

If user authentication information is replicated, they should be stored at sites where the users access the system.

Distributed authorization rules are stored in the catalogue. Handling user groups for authentication simplifies distributed database administration.

Full replication of authorization has 2 advantages:
– Simpler
– Can be done at compile time

However disadvantage is that there is a higher overhead cost.

Distributed Semantic Integrity Control

The main problems in designing an integrity subsystem for a distributed DBMS are:
– Definition and storage of assertions
– Enforcement of these assertions

Definition of Distributed Integrity Assertions

3 classes of assertions:
– Individual assertions: single-relation single-variable assertions.
– Set-oriented assertions: single-relation multivariable constraints
– Assertions involving aggregates: requires special processing due to cost of evaluating the aggregates.
New integrity assertion: Sites that store the relations involved in the assertion.

An integrity assertion becomes a distributed operation in 2 steps:
– Transform high-level assertions into compiled assertions
– Store compiled assertions according to class of assertion.

Enforcement of Distributed Integrity Assertions

Main problem is to decide at which site to enforce the integrity assertions.

Choice depends on:
– Class of assertion
– Type of update
– Nature of site where update is issued.


“————————————————– ————————————————– ————————————————–

Appoġġ traduzzjoni: http://amzn.to/1Z7d5oc
————————————————– ————————————————– ————————————————–

Kontroll tad-Dejta Semantika

kontroll tad-data semantiku tipikament tinkludi l-ġestjoni fehma, kontroll tas-sigurtà u l-kontroll integrità semantika. Dawn il-funzjonijiet jiżguraw li l-utenti awtorizzati jwettqu operazzjonijiet korretta fuq il-bażi tad-dejta. Dan jikkontribwixxi għall-integrità database.

Fehmiet fil DBMS Imqassam

Veduta f’database mqassma ma tkunx ġejja mid-relazzjonijiet frammentati maħżuna f’siti differenti. Meta ħsieb huwa definit, l-isem tagħha u mistoqsija irkupru tagħha huma maħżuna fil-katalogu.

L-immappjar ta ‘mistoqsija espress fuq opinjonijiet isir bl-istess mod bħal fis-sistemi ċentralizzati. Fil sistemi ċentralizzati dan isir permezz ta ‘modifika mistoqsija. A mistoqsija mqassma hija pproċessata mill-proċessur mistoqsija mqassma.

Fehmiet fi DBMS distribwiti jista ‘jsir bl-użu regoli simili għar-regoli ta’ definizzjoni framment. Il-proċessur mistoqsija mapep tal-mistoqsija mqassma fi query fuq frammenti fiżiċi.

Opinjonijiet miksuba mid- database mqassma tista ‘tkun għalja biex jevalwa. View derivazzjoni isir billi jingħaqdu il-kwalifika fehma mal-kwalifika mistoqsija. L-alternattiva hija li tuża Snapshots. Daqqa t’għajn lejn jirrappreżenta stat partikolari ta ‘database. Dan ifisser li huwa statiku. Snapshots huma utli meta l-utenti ma jkunux interessati li jaraw l-aktar verżjoni reċenti tad-database. Aċċess permezz Snapshots huwa adegwat għall-mistoqsijiet li huma stabbiliti minn qabel.

Kontroll Awtorizzazzjoni Imqassam

Aktar kumplessi minn f’databases ċentrali minħabba l-fatt li l-oġġetti u l-utenti jistgħu jitqassmu. Dawn il-problemi huma:
– Awtentikazzjoni utent Remote
– Ġestjoni tar-regoli ta ‘awtorizzazzjoni mqassma
– Żamma ta ‘fehmiet u gruppi ta’ utenti.

Biex jipprevjenu l-aċċess utent remot mill-utenti mhux awtorizzati, 2 approċċi huma użati:
– Informazzjoni ta ‘utenti awtorizzati huma maħżuna fuq is-siti kollha fil-katalgu (iktar għaljin)
– Is-siti kollha ta ‘DBMS distribwiti jawtentikaw lilhom infushom bħall-utenti do. (Meħtieġ jekk l-informazzjoni l-utent ma jkunx replikati).

Jekk l-informazzjoni awtentikazzjoni utent huwa replikat, dawn għandhom jiġu maħżuna f’siti fejn l-utenti jaċċessaw is-sistema.

regoli ta ‘awtorizzazzjoni distribwiti huma maħżuna fil-katalgu. Handling gruppi ta ‘utenti għall-awtentikazzjoni jissimplifika l-amministrazzjoni dejtabejż mqassma.

replikazzjoni sħiħa tal-awtorizzazzjoni għandu 2 vantaġġi:
– sempliċi
– Jista ‘jsir fi żmien jikkompilaw

Madankollu iżvantaġġ huwa li hemm spiża overhead ogħla.

Imqassam Kontroll Integrità semantika

Il-problemi ewlenin fit-tfassil ta ‘subsistema integrità għal DBMS distribwiti huma:
– Definizzjoni u ħażna ta ‘affermazzjonijiet
– Infurzar ta ‘dawn l-affermazzjonijiet

Definizzjoni-affermazzjonijiet Integrità Imqassam

3 klassijiet ta ‘affermazzjonijiet:
– Affermazzjonijiet individwali:-relazzjoni unika affermazzjonijiet wieħed-varjabbli.
– Orjentata lejn is-Set affermazzjonijiet: restrizzjonijiet-relazzjoni unika multivariable
– Dikjarazzjonijiet li jinvolvu l-aggregati: jirrikjedi speċjali ta ‘ipproċessar minħabba l-ispiża ta’ evalwazzjoni tal-aggregati.
Ġodda affermazzjoni integrità: Siti li jaħżnu r-relazzjonijiet involuti fl-affermazzjoni.

Affermazzjoni integrità ssir operazzjoni distribwiti fi 2 passi:
– Transform allegazzjonijiet ta ‘livell għoli fis affermazzjonijiet kkumpilata
– Aħżen ikkumpilata allegazzjonijiet skond il-klassi ta ‘affermazzjoni.

Infurzar tal Dikjarazzjonijiet Integrità Imqassam

problema prinċipali hija li tiddeċiedi fuq liema sit biex tinforza l-affermazzjonijiet integrità.

Għażla tiddependi fuq:
– Klassi ta ‘affermazzjoni
– Tip ta ‘Aġġornament
– In-natura tas-sit fejn aġġornament jinħareġ.


“————————————————– ————————————————– ————————————————–

Ondersteuning vertaling: http://amzn.to/1Z7d5oc
————————————————– ————————————————– ————————————————–

Semantiese Data Control

Semantiese data beheer sluit tipies oog bestuur, sekuriteit beheer en semantiese integriteit beheer. Hierdie funksies verseker dat gemagtigde gebruikers uit te voer korrek bedrywighede op die databasis. Dit dra by tot die databasis integriteit.

Uitsig in Distributed DBMS

‘N Blik in ‘n verspreide databasis is afgelei van gefragmenteerde verhoudings gestoor op verskillende plekke. Wanneer die oog word gedefinieer, sy naam en sy herwinning navraag word gestoor in ‘n katalogus.